Single Cell Mass Cytometry of Non-Small Cell Lung Cancer Cells Reveals Complexity of In vivo And Three-Dimensional Models over the Petri-dish
Received: 31 July 2019 / Revised: 12 September 2019 / Accepted: 15 September 2019 / Published: 16 September 2019Abstract
Single cell genomics and proteomics with the combination of
innovative three-dimensional (3D) cell culture techniques can open new
avenues toward the understanding of intra-tumor heterogeneity. Here, we
characterize lung cancer markers using single cell mass cytometry to
compare different in vitro cell culturing methods: two-dimensional (2D),
carrier-free, or bead-based 3D culturing with in vivo xenografts.
Proliferation, viability, and cell cycle phase distribution has been
investigated. Gene expression analysis enabled the selection of markers
that were overexpressed: TMEM45A, SLC16A3, CD66, SLC2A1, CA9, CD24, or repressed: EGFR
either in vivo or in long-term 3D cultures. Additionally, TRA-1-60,
pan-keratins, CD326, Galectin-3, and CD274, markers with known clinical
significance have been investigated at single cell resolution. The
described twelve markers convincingly highlighted a unique pattern
reflecting intra-tumor heterogeneity of 3D samples and in vivo A549 lung
cancer cells. In 3D systems CA9, CD24, and EGFR showed higher
expression than in vivo. Multidimensional single cell proteome
profiling revealed that 3D cultures represent a transition from 2D to in
vivo conditions by intermediate marker expression of TRA-1-60, TMEM45A,
pan-keratin, CD326, MCT4, Gal-3, CD66, GLUT1, and CD274. Therefore, 3D
cultures of NSCLC cells bearing more putative cancer targets should be
used in drug screening as the preferred technique rather than the
Petri-dish.
Introduction
Development of single-cell analytical techniques extends our understanding of cell population heterogeneity and enables the identification and characterization of highly specialized rare cell types . Single cell genomics (e.g., single cell RNAseq) and single cell proteomics (e.g., mass cytometry) have revolutionized our knowledge about the co-ordination of different cell types in tissue microenvironments unveiling their characteristic protein patterns . Although image-based single cell analysis has also been developed single cell mass cytometry has been adopted to investigate millions of cells per sample and it offers multi-dimensional data analysis with the characterization of multiple proteins at single cell resolution. Lung cancer accounts for the majority, 25% of all cancer-related deaths worldwide and the 5years overall survival at 17.7% has achieved very little progress in the last decades . Adenocarcinomas account for the majority, 40% of all lung cancer histological types . Here, we focus on the mass cytometric single cell analysis of a non-small cell lung carcinoma (NSCLC) model, the A549 adenocarcinoma cells. The heterogeneity of immune subsets infiltrating non-small cell lung cancer has been previously published based on mass cytometric profiling . Here, we focus on marker expression of lung cancer cells obtained from different culture conditions at single cell resolution. Organoid cell culturing revolutionized cell biology since in vitro three-dimensional (3D) multicellular spheroid models mimic better the physiology of complex tissues compared to conventional two-dimensional (2D) monolayer cultures . For a more successful treatment of lung cancer, a better understanding of cancer development in the tissue microenvironment and further improvements in in vitro experimental techniques are needed. Ideal models systems, for screening of novel drug candidates should mimic the molecular, functional and histopathological complexity of in vivo tumors more accurately. Although multi-cellular tumor spheroid models were introduced in the early 1970s , their implementation in the pre-clinical phase of drug development was neglected resulting in numerous failed clinical trials .
Materials and Methods
2.7. Imaging Digital phase contrast images were taken by the HoloMonitor M3 instrument using phase contrast X10 objective (Phase Holographic Imaging AB, Phiab, Sweden) and the analysis computer (HoloStudio 2.0 software, Phiab, Sweden). Phase contrast images were used as a reference to confirm that the cells were in good condition under the studied period. Changes were analyzed at days 4 and 9.
 |
| Figure 1. Different culture conditions of A549 human adenocarcinoma cells for 4 days (A) and for 9 days (B). Cells were seeded at the same surface/volume ratio regarding the different culture conditions as described in Materials and Methods. Images were taken by the HoloMonitor M3 instrument using phase contrast X10 objective. Scale bar: 150 µm. |
Discussion
Genomics and proteomics opened new avenues for drug discovery with the emergence of novel therapeutic and diagnostic targets. Recent achievements in drug discovery with the combination of innovative 3D cell culture techniques yielded high-throughput screening (HTS) methodologies of 3D cellular assays in the pre-clinical phase of the drug discovery pipeline. These 3D HTS assays provide information not only on a general cellular response (cytotoxicity) to a given drug, but could map a signal transduction machinery or monitor the cellular response at transcriptional/translational level in a model system closer to the in vivo situation . Monolayer cultures are oversimplified models of a multicellular organism. Beyond their benefits, they fail to present the in vivo situation in several aspects such as the flow of cellular metabolites, partial oxygen tension, the gradient of soluble mediators, whereas 3D tissue culture models have a closer resemblance to the in vivo conditions .
In 3D models oxygen and nutrient deprivation results in higher glycolytic activity, elevated autophagy and necrosis induced by anaerobic conditions . Active signaling pathways are also different between 3D and 2D cultures for the lack of complexity of cell-cell and cell-ECM connections (integrins, proteoglycans) present in multicellular organism . Different studies show that multicellular spheroid formation highly influenced extracellular matrix protein expression and relevant gene expression levels as well . In order to find the most suitable 3D culture method to mimic in vivo tumor biology we performed comparative analysis of different 2D (2D standard Petri-dish; tissue culture-treated T-75 flasks, 2D TC; type I collagen coated plates, 2D Coll); 3D bead-based (3D Cytodex3, 3D Nutrisphere); 3D carrier-free (3D Spheroid) and collagen embedding (RAFT) cell culture methods . These culture methods were compared to A549 xenograft tumors (in vivo).
Both short-term (day 4) and long-term (day 9) cultures were analyzed with early stage and late stage adenocarcinoma. The investigation of the proliferation rate in different culture methods revealed that carrier-free spheroids divided less frequently with the longest lag-phase, and viability was hampered in longterm RAFT cultures. These two 3D culture methods were excluded from our single cell experiments, due to their difficult handling and standardization of spheroid size. On the contrary, bead-based systems offer a constant and calculated surface/volume ratio. Both 3D Cytodex3 and Nutrisphere beads can be counted using standard methods, therefore the density of beads can be controlled for each experiment. Cell cycle phase distribution showed moderate changes in only shortterm cultures (day 4), but not in long-term (day 9) with significant enrichment in the G2/M phase in 2D Coll, 3D Cytodex3 and 3D Nutrisphere cultures (Figure 3).
Gene expression analysis by nanocapillary qRT-PCR (624 genes) and 1536 well high-throughput qRT-PCR (62 genes) resulted in the selection of lung cancer markers associated with higher (TMEM45A, SLC16A3, CD66, SLC2A1, CA9, CD24) or lower (EGFR) expression in vivo or in 3D models compared to monolayer cultures (Table 2, Figure 4–5). Additionally, TRA-1-60, pan-keratins, CD326, Galectin-3, CD274 with known clinical significance were also included into the panel. Adenocarcinomas express keratins (K) such as: K8, K18, K19 and in some cases also K7 and K20 as building blocks of intermediate filaments . Interestingly, these cytokeratins also localize to the cell surface of carcinoma cells to enhance adherence to ECM . Therefore, an anti- pan-keratin antibody (clone C11) was used for cell surface labeling which recognizes keratins 4, 5, 6, 8, 10, 13 and 18 . The implemented multidimensional single cell proteome profiling revealed that 3D (Cytodex3 and Nutrisphere) cultures represent a transition from 2D to in vivo situation by intermediate marker expression of TRA-1-60, TMEM45A, pan-keratin, CD326, MCT4, Gal-3, CD66, GLUT1, CD274. In 3D systems CA9, CD24, EGFR showed higher expression than in vivo (Figure 6).
Our multi-parametric single cell mass cytometry results delineated a map with different regions that represented 2D, 3D and in vivo conditions with minimal overlap (Figure 7). Our single cell study was able to detect the rate of heterogeneity in 2D, 3D cultures and in solid-tumor (Figure 6–7). As a result, the following proteins were associated with advanced cancer: TRA-1-60 (9%), TMEM45A (37.5%), pan-keratin (50%), CD326 (100%), MCT4 (50%), GAL-3 (30%), CD66 (100%), GLUT1 (100%) (Figure 8A).
The complexity of an organism is far from cell culture systems, but bead-based 3D cultures provide a better representation of the in vivo conditions affording a more effective methodology for different molecular biology studies, as well as for screening of compound libraries for novel anticancer drug candidates.
Inga kommentarer:
Skicka en kommentar